In article <[EMAIL PROTECTED]>, Julien ALLANOS <[EMAIL PROTECTED]> writes:
> Thanks, I've already read these pages. > Here is a test example I've done: > 1/ User A: SET GLOBAL TRANSACTION ISOLATION LEVEL READ COMMITTED; > 2/ User B: SET GLOBAL TRANSACTION ISOLATION LEVEL READ COMMITTED; > 3/ User A: START TRANSACTION; > 4/ User B: START TRANSACTION; > 5/ User A: SELECT * FROM MyTable LIMIT = 1; > I got a one-row result. > 6/ User B: SELECT * FROM MyTable LIMIT = 1; > I got the same one-row result, and I need to get the next one instead, because > this one is already being treated by user A! > I really want to have a "SELECT FOR DELETE" thing: as soon as a row is read, > it > is deleted. Is this possible? Not quite, but close. You need a SELECT.. FOR UPDATE. Then in step 6/ above the B connection will get blocked. Now you can delete the row in the A connection and commit. Then B will get unblocked and return the "next" row. Just two things: 1. Use "ORDER BY pkey LIMIT 1" instead of just "LIMIT 1". You might get away without the ORDER, but the behavior is undefined. 2. Remember the result of the "SELECT.. FRO UPDATE", do the DELETE and COMMIT immediately, and process the row values afterwards if possible. This will cut down the time B gets blocked. -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
